org.tribuo.classification.sgd.objectives

Class LogMulticlass

java.lang.Object

org.tribuo.classification.sgd.objectives.LogMulticlass

All Implemented Interfaces:

com.oracle.labs.mlrg.olcut.config.Configurable, com.oracle.labs.mlrg.olcut.provenance.Provenancable<com.oracle.labs.mlrg.olcut.provenance.ConfiguredObjectProvenance>, LabelObjective

public class LogMulticlass
extends Object
implements LabelObjective

A multiclass version of the log loss.

Generates a probabilistic model, and uses an ExpNormalizer.

Constructor Summary

Constructors

Constructor and Description
LogMulticlass()

Method Summary

Modifier and Type	Method and Description
VectorNormalizer	getNormalizer() Generates a new VectorNormalizer which normalizes the predictions into [0,1].
com.oracle.labs.mlrg.olcut.provenance.ConfiguredObjectProvenance	getProvenance()
boolean	isProbabilistic() Returns true.
String	toString()
com.oracle.labs.mlrg.olcut.util.Pair<Double,SGDVector>	valueAndGradient(int truth, SGDVector prediction) Returns a Pair of Double and the supplied prediction vector.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface com.oracle.labs.mlrg.olcut.config.Configurable

postConfig

Constructor Detail

LogMulticlass

public LogMulticlass()

Method Detail

valueAndGradient

public com.oracle.labs.mlrg.olcut.util.Pair<Double,SGDVector> valueAndGradient(int truth,
                                                                               SGDVector prediction)

Returns a Pair of Double and the supplied prediction vector.

The prediction vector is transformed to produce the per label gradient.

Specified by:

valueAndGradient in interface LabelObjective

Parameters:

truth - The true label id

prediction - The prediction for each label id

Returns:

A Pair of the score and per label gradient.

getNormalizer

public VectorNormalizer getNormalizer()

Description copied from interface: LabelObjective

Generates a new VectorNormalizer which normalizes the predictions into [0,1].

Specified by:

getNormalizer in interface LabelObjective

Returns:

The vector normalizer for this objective.

isProbabilistic

public boolean isProbabilistic()

Returns true.

Specified by:

isProbabilistic in interface LabelObjective

Returns:

True.

toString

public String toString()

Overrides:

toString in class Object

getProvenance

public com.oracle.labs.mlrg.olcut.provenance.ConfiguredObjectProvenance getProvenance()

Specified by:

getProvenance in interface com.oracle.labs.mlrg.olcut.provenance.Provenancable<com.oracle.labs.mlrg.olcut.provenance.ConfiguredObjectProvenance>